Artificial Intelligence: How tech tools are changing work, school and the stock market

AI has become a part of everyday live for Americans, whether we use the technology directly or not. AI-driven productivity and communications tools dominate the online experience. Speculation about the future of a workforce displaced by AI technology is rampant. And massive data centers built to house generative AI infrastructure are popping up nationwide, often met with opposition by local communities. Two journalists who’ve reported on the impact of AI in their respective beats, Justin Lahart of the Wall Street Journal and Matt Barnum of Chalkbeat Ideas, join us to break down these talking points.

Linda O’Bryon: Artificial Intelligence, a term first coined for a 1956 conference in Hanover, New Hampshire. The goal? Describing every aspect of learning or other features of intelligence so a machine can be made to simulate it. By thinking of ways to make machines more cognizant, participants of the Dartmouth Summer Research Project on artificial intelligence wanted to lay the framework to have a better understanding of human intelligence. I'm Linda O'Bryon.

Nicole Cox: And I'm Nicole Cox. Flash forward 70 years to find a better understanding of human intelligence may not be as front of mind for the general public. The real-world demands of energy, water, chips, and hardware for data centers feels voracious to some and is resulting in increased costs for Americans that go beyond electricity and unease over how AI will redefine jobs.

LO: The value of tech giants and AI-linked firms has been climbing for the past few years and despite a booming stock market, people watching the market are concerned. Tech companies are borrowing billions of dollars for the construction of data centers, purchasing chips and hardware, and to acquire top talent.

Recent skepticism over future earnings, coupled with concerns we are in an AI bubble, has caused volatility and even warnings from the International Monetary Fund of what might happen when the bubble bursts.

NC: Our promises of longer term gains enough for the digital market are not just a negative for short and medium financial and environmental pains? Plus growing worries that AI will take over more and more jobs.

Joining us to discuss is Justin Lahart, reporter with The Wall Street Journal. Thank you so much for joining us today!

Justin Lahart: Thanks for having me.

LO: There has been some volatility in top AI stocks. The Wall Street Journal, your publication reports about eight to 10% of government pension funds are invested in a basket of publicly traded AI-related companies. Meta, Alphabet, Microsoft, and Amazon spent $130 billion in the first quarter of 2026. Based on that spending, AI-related investments could go as high as $725 billion for the year as a whole.

Justin, what has caused the recent volatility in the AI stocks and where do you think we stand with AI as an investment?

JL: Yeah, when I think about the volatility, there's always so many things going on. One thing about these, you know, the so-called hyperscalers—Amazon, Microsoft, Meta, and so forth—they have substantial capital investment needs. Analysts expect them to invest something like $4 trillion over the next four years. That's a lot of money. So they're gonna have to raise money to be able to pay for all that. A lot of that will come through the bond market with a lot of uncertainty on where rates are, that causes some concern on its own, and then you're just looking at an enormous amount of money. You know, what's the payoff gonna be?

You know, the other thing about AI is, the payoff isn't gonna come tomorrow. The payoff is gonna come a long, long time from now, right? Four, five, six, seven years out. It's a long-term investment. Long-term investments like that are very, very sensitive to interest rates. So again, we have sort of concerns about what is the rate situation gonna be? What's inflation going to look like? It can really, really affect those kinds of companies. So just that can create a lot of volatility within these very, very big, very important to people's retirement fund stocks.

LO: Because if interest rates continue going up, their borrowing costs and investment costs are also gonna go up.

JL: Right, exactly. Yeah, and also you sort of, you know, you wonder about the payoff. The payoff has to be even more to sort of justify the price that people are paying for the stock.

NC: So keeping that long-term payoff in focus then, from an investment standpoint, what signals would you look for to determine if AI-related companies will continue to be the darling of Wall Street, or whether they're reaching bubble territory? Does that necessarily mean a major retreat? Does the future of AI mean for the overall stock market something different and for the economy at large because of that long-term payoff you mentioned?

JL: Yeah, so bubbles are hard. The best sort of definition I've ever heard of a bubble is that people pay a crazy price and they trade like crazy. That is definitely happening. You just don't know where the end is. Is it a bubble yet? Will it be a bubble on Wall Street?

You know, I think a lot of people do think it's a bubble, and they also sort of assume that they're gonna be able to get out in time. Everyone thinks they'll make their money and then all get out in time. So, it does just make you concerned about the stock market and just the overall valuations of the stock market are incredibly high historically. So there's a lot to worry about there. Again, people just don't wanna miss out.

LO: Well, but on the other hand, there's this voracious demand for AI, for the product itself, for the services, everything that it does. So if you think of it that way, is this just the beginning of something even much greater?

JH: I think the prices that we are paying for AI now are probably much lower than the actual cost is. Kind of a loss leader now, and eventually either the costs to the companies will fall a lot or they'll charge us a lot more. So boy, it sure is confusing.

LO: Well, in the meantime, the data centers that are being produced to support AI are having an effect on the overall economy. For example, data from the US Consumer Price Index indicates that electricity prices are up 4% from a year ago. And for the first five months of this year, those prices rose even faster than 2025. Another example: Apple increased prices for popular products by 30%, saying AI data centers created an extraordinary surge in demand for memory and storage. So growing demand for computer chips could drive up prices for medical devices, vehicles, even kitchen appliances. What are you hearing about how consumers are holding up with these price increases?

JL: You know, it's gonna be very interesting with what happens with electricity, right? It's not going to drive up like, you know, different people live in different places and what AI, what these data centers could do to their electricity bills will vary, right? People in Texas where a lot of data centers are being built, they could really see their electricity prices go up a lot. You know, the same for Virginia, maybe some other places where these data centers aren't being built, aren't gonna have the same effect.

Overall, I mean, economists don't think it's going to really push up electricity prices all that much overall, but, right, some people in some places could really see it and they'll notice it. And then there's other concerns about what happens with electricity, right? These data centers could put a lot of pressure on the grid. So it might not raise your electricity prices all that much, but during a heat wave where there's not a lot of buffer anymore for electricity, all of a sudden you could have brownouts and blackouts and problems like that. That will be a cost. It will not show up any price, but that will be a cost that people will pay.

Another thing is, you know, as these data centers roll out, you could have a lot more reliance on gas-fired power plants. So that could make electricity prices a lot more responsive to changes in natural gas prices, for example. So you could have a lot more volatility. So there's just a lot of moving parts there when it comes to electricity. And it might not all show up in the price, right? But it will show up for people. I think when it comes to things like the price of the iPhone going up a lot, the price of various game controllers going up a lot because of these chip prices. People don't buy a phone every day. They don't buy a game controller every day. It's not a big part of our consumer basket, but it is very salient. It's something that we really, really notice. People will start shopping for Christmas and stuff like that. And they notice that Nintendo game controller is up a lot. They're gonna notice it. And they're gonna connect the dots and say, oh, this has something to do with AI. And that could tick people off.

LO: And we're starting to see some of that already. We'll talk more about that in a moment, but you have written extensively about economists and how they view the impact of AI. From your observation, what would you say the single biggest factor to the economy is as a result of AI so far?

JL: I mean, so far, the biggest factor has just been all of this money that is being pumped into AI. So everybody is thinking, well, what's it gonna do to productivity? Is it gonna make us more efficient and more productive? Economists generally think, yes, it will, but not yet. Will it affect the job market in good or bad ways? It's not really, it's maybe showing up on the margins, but not really yet. The biggest effect so far has been what it's doing to prices and that there is just this gusher of money that is going into AI investment. And that money is not being spent on other things. It's really supporting the economy. On the other hand, there are other things that that money might've gone into that just isn't happening now. So it's really just kind of changing things that way.

NC: What about the local versus national economies? Are we seeing the impact is pretty evenly spread on both?

JL: It has to be pretty varied. I mean, for sure, right? There are people and there are Silicon Valley workers who are really benefiting from AI. People who are working for these companies and suddenly doing extremely well and that affects those economies. Definitely in some of these areas where these data centers are being put in place, there are pretty profound effects. People being bought out of their land for just millions of dollars and there's been a lot of pushback against that. It's having these really very localized effects. I think when it comes to sort of AI use, AI use in the workforce, it's probably not moving the needle as much as we might think it is.

NC: I was interested to hear that you said AI is a loss leader. And of course, the first thing that popped in my mind is how will that change as people become more dependent upon it?

JL: You would think, right, computing in the past, right? Technology, it's expensive and then it gets cheaper to produce. So is that something that happens with AI? To the point where we are cheap enough to produce that these companies are making a profit off of it or we've become accustomed to it enough that it's worth it for consumers and for companies to use it. Is it ultimately something that was really only cost-effective for people doing very high level research, right? That could be a possibility. And there's just so many questions.

LO: You mentioned the impact in local communities. There's also a global impact, which is interesting. According to the Rice University's Baker Institute for Public Policy, there has been a decline in total imports and a shift in trade volumes due to tariff policy changes. We've seen a lot of tariffs have gone up and down.

But as the AI infrastructure boom happens, we're seeing all these data centers. And also this huge need for chips. So is in a way the demand for AI shifting the global trade economy as we know it? Where certain countries like Taiwan, where they have chips, are more of a trading partner or becoming more of a trading partner with the United States than say Mexico?

JL: I mean, Taiwan is definitely exploding, right? Their exports to the US are exploding. Interestingly enough, Mexico too. They export a lot of tech products that are going into AI. So that is really happening. I guess it's probably countermanding some of the effects of the tariffs. So then the prices are just amazing, right? Because these countries, these companies just can't produce this stuff fast enough. I saw in July import prices for computers and semiconductors. It was up 17% versus a year ago. I mean, that's a huge, huge number. Usually when we're looking at computer and chip prices, versus a year earlier, they're going down, right? They get better and better and better. And the way we measure it is that better chip.

NC: Yeah, production becomes more efficient. No, but you're absolutely right with production becoming more efficient. We're expecting these prices to go down, but now we're seeing the opposite because of demand and also how that's impacting other aspects of the economy. CNN Business reports, pricing pressures are most evident at the producer level with the producer price index chart, semiconductor and other electrical component manufacturing industry looking like a hockey stick. However, the new feature co-pilot, Microsoft's new AI tool. Certainly it will be interesting to see if any of these features become somewhat of an a la carte feature, just as what happened with cable or satellite or how people get their streaming service. But I thought that was very interesting. Linda?

LO: So we talked earlier about some people who are pushing back on AI. And there was an example in July when a hundred people gathered outside OpenAI's Mission Bay offices in San Francisco to protest this rapid development of artificial intelligence by Bay Area companies. They had chants saying, "Slam the brakes, slow the race."

America's federal government, state as well, both appear to have left the technology in the hands of the company, the regulations. Do you see that changing as we keep seeing more and more concerns about AI and various elements of deep fakes, of misinformation, ways that it's being misused?

JL: The opinions about AI are really, have become quite negative. You might've seen, so Pew was out recently with some polling that showed just the number of people who are more concerned about AI, vastly outnumbers the number of people who are excited about it.

And interestingly, this is especially true for young people. I would normally expect young people to be jazzed about a new technology. When the internet was new, young people were really, really excited about it. Maybe older people weren't as excited about it. But in this case, when you talk to people in their 20s, they really don't love it. They're really concerned about what it's gonna do to the job market. If you talk with college students, they don't really like what it's doing to education, to educational process, because they feel like they're competing with students who are using AI, so I have to use AI too.

There's just all of this pushback. Then there's concerns about deep fakes, privacy. It is a situation where you could really see pushing for more regulation happening.

LO: Well, maybe, could we go back to that whole notion of government regulation, because there is pushback. Do you think government regulation is gonna come from the state and local level, or will Congress step in?

JL: Well, I mean, we're already seeing it at the state and local level, right? With states like New York putting a pause on new data center construction and so forth. It seems like the type of thing that might start at the state and local level, and then push its way up to the national level. I suspect that in the upcoming elections, for example, that we might really see some moves at state and local, and that it might not really move the needle on what's going on with Congress.

However, two years from now, during the next presidential election cycle, that situation could be very different, right? It could be part of the platform. The pushback could be a lot more extreme, depending on where things go. I guess it's sort of a stay-tuned situation, I don't know.

LO: Well, one of the things that could really push it is something that has happened just in the last few months. We've heard reports of Meta’s AI models going rogue in terms of their testing, hacking third-party devices. OpenAI saying some of its experimental AI models hacked into another company's production system. I've heard people talking about these AI agents talking to each other and bypassing humans.

So if we hear more of these kinds of examples, doesn't that suggest that someone other than the companies themselves are gonna have to come in and regulate?

JL: You would think that that's it. If you have more of those examples, you'll have, there'll be more of a regulatory push. I think that just the pushback against data centers might be even more salient for people. They really do not want a data center in their backyard. And even we're seeing in some of these rural communities where people being offered really amazing amounts of money and turning it down because it affects their way of life. It's the way of life that they're used to. They don't want to live right near a data center. That's not what they're accustomed to. That's not what they consider part of their quality of life and money doesn't solve that situation for them.

NC: Could you explain why economists are concerned over the way AI is being designed by these companies?

JL: A lot of what the companies have pushed, they've sort of gone back on this recently because of the criticism, but a lot of the sort of thought on designing AI is to use it to replace workers. ‘I'm gonna design an AI system and it's gonna replace all the paralegals or all the accountants,’ or something like that. That's the design is to take tasks that humans do and do those tasks better and then you don't need the humans anymore, which is great for the bottom line of a company that you might be selling AI to.

It's a very different sort of way of thinking about technology than the personal computer was. So famously, Steve Jobs read a piece in Scientific American about how efficient different animals were and the most efficient sort of animal on this chart that Scientific American had was a person on a bicycle. And he came up with this idea that the computer should be a bicycle for the mind.

That was really what motivated tech companies for a long time, that we were gonna create technology that was going to augment what we could do, that would make us more productive, that would make us be able to do more, not technology that would replace us. And AI is being more designed as a replacement technology. That might not be a very sort of fruitful direction for a technology. It might be better to take a page out of the old book and start thinking about using it as a technology that would make you perform your job better, make us more productive, make us have better lives, produce more, not something that is gonna sort of lay us off and then somehow the money will be redistributed to us while we don't work.

LO: Who could make that happen?

JL: Who could make that happen? I mean, I think that some of it might be regulatory, some of it might be just a shift in the demands that people make of AI. And to be sure, could be that as people use AI, that it does turn out to be more of an augmentative technology that ultimately we find that its best use is not to replace a worker, but to allow people to do more and more. It could be that sort of existing companies are gonna use AI to try and reduce their workforce and that there will be new companies that come along and they'll be using AI in a different way and saying, well, no, we're just gonna do more and have better, more productive workers and maybe those newer companies, we'll end up eating the old company's lunch.

LO: Fascinating. On that hopeful note, thank you very much, Justin Lahart. We appreciate your time and Justin's reporting can be found at wsj.com. Justin Lahart, thank you for joining us today.

LO: For years, American industry has influenced what's been taught in the classroom. When we come back, Matt Barnum with Chalkbeat discusses the role of big tech and generative AI in schools and how the economy might adapt.

You're listening to Navigating Our Financial Times.

NC: Welcome back to Navigating Our Financial Times. I'm Nicole Cox.

LO: And I'm Linda O’Bryon. We are now joined by Matt Barnum, editor and columnist for Chalkbeat Ideas. Thank you for joining us today, Matt.

Matt Barnum: Happy to be back. Thanks for having me.

LO: Well, whether working to ensure that factory workers have basic literacy and math skills or hands-on vocational training, American industry’s influence over public education is not new.

NC: And for years, the economy adapted to new technologies by requiring more education. Matt, your article titled, For generations, more schooling meant better jobs. Could Generative AI Upend That Bargain?, it looks at the interplay between technology, the economy, and schooling. Could you give us an overview of how technology has historically transformed education?

MB: For a century, technology has made working with your head more and more valuable than working with your hands. That's not a values judgment. That's just a fact of the modern economy. It used to be that huge swaths of Americans say, worked in farms or worked in manufacturing. But that isn't the case as much anymore. And in fact, the people who earn the most money tend to be those who got college degrees or even more, lawyers, doctors, accountants. And that has led to many more people going to college because your best chance for many people of earning a good living is getting a college degree because that is what is rewarded in the modern labor market with modern technology that can do stuff that we used to do with more people working in factories or more people working in farms. And that has just led more and more emphasis on schooling and education.

LO: Is that changing dramatically?

MB: The short answer is we don't know yet. And I think it's fair to say that it hasn't changed dramatically yet, but we are very, very early in artificial intelligence. And it's at least more plausible that it could change. People have talked about a white-collar job apocalypse because artificial intelligence can do a lot of things that white collar workers are doing, myself included, by the way. It can write, it can analyze, it can synthesize. And for those of us who have jobs that do that, that are educated to do that, AI, it's at least more plausible than prior technologies, like improved farm equipment, that AI could displace white collar workers.

But I don't think that's a guarantee. There was recently, the Wall Street Journal surveyed about 10 leading economists recently, and they asked them, will AI lead to more white collar jobs about the same or fewer? And they were almost evenly split. There were some who said it would mean more white-collar jobs, some who said that there would be fewer and others who said it would be about the same. I just think it's more plausible now than it's ever been.

LO: And ironically, AI is fueling so much of the stock market, which then in turn affects the economy. But we've also seen the nation's debt increasing. August 18th: reports that the debt had exceeded $40 trillion, more than double its level in 2017. Analysts say that rising federal debt could affect the country's ability to handle future economic shocks, including AI disruption or a recession. If people become wary and investors start to pull money out of the US believing that an AI bubble is about to burst, what impact will that have on education? I mean, of course, education has value beyond its economic return, but education is a key driver in this economy.

MB: If AI were to change the economy in a way that devalued education, then our current structure of schooling would dramatically shift. We set kids up to go through K-12 education and for large swaths, the majority to continue their education. And the K-12 schools, our public schools and also private schools, are set up to prepare many students in basic literacy and numeracy and more advanced literacy and numeracy to get ready for college.

So if AI changed that bargain where there isn't this big economic return to education, it would really throw the entire structure of American education from kindergarten, which by the way has become more sort of academically focused than it used to be, from kindergarten to colleges and universities.

NC: You referenced that bargain. You do write about the knowledge economy, the reason each generation of Americans remain in school longer than the last since workers with the most formal schooling, as you mentioned earlier, tend to be the best paid. Economists refer to this as the race between education and technology.

Why is generative AI so different and how might the economy respond?

MB: And I would say it could be different, but we don't know. And I think that's really important that we're just, there's a lot of uncertainty, but we don't know about the basic facts about what's happening right now at this very, very moment.

My basic answer to your question is we still don't know how this is going to play out. A lot of technologies, there's been a lot of predictions, for instance, that there was one example recently of an analysis of like technology completely replacing the bank teller. That didn't happen at the rate that people thought, though now a different form of technology, as I understand it, the mobile phone and people being able to make bank deposits on their phone, has led to reduction in bank tellers, but that was not the predicted route.

And so I just think there was a lot of uncertainty right now. And we should, as journalists, we should keep our minds open and we should follow the data to understand what's happening.

LO: Well, in the meantime, AI companies are an employer, a pretty big employer and growing. And some of them are hiring mathematicians, physicists, computer science professors, economists, top researchers from universities. Anthropic has a team of philosophers. And we've even heard of an AI company recruiting a scientist who studies black holes and another who specializes in string theory. So these are different kinds of careers. What's behind this shift and what could it signify for the demand for certain careers?

MB: I don't know what is driving that because I'm not exactly sure why the companies have done that. I think they have their own interests and needs and they're growing rapidly and they're trying to get a lot of talented people with a lot of different backgrounds, a lot of different skillsets. And I think they're also seeing a lot of uncertainty. And so they want a variety of skillsets. It strikes me as this could be a drop in the bucket though, relative to America's colleges and universities, which employ large swaths of mathematicians and others, and philosophers and others who have PhDs. So if the bargain of education falls apart and American universities are less valued than they used to be, I'm kind of skeptical that AI companies would be able to fill the gap by hiring philosophers and mathematicians, but we'll see.

NC: As the debate continues over AI in schools, AI-powered private schools, Alpha Schools, are expected to grow to about 50 campuses across the country this year. The Texas-based AI school model has students using AI software for core subjects. There are no licensed teachers but guides who are motivating and coaching the students. Is there a good argument for schools to embrace AI or is this an example of education operating as an economic enterprise?

MB: We need to understand the results of these schools in more detail, which are private schools and where data is often hard to compare to students in public schools. As I understand it, the vast majority of Alpha Schools charge tuition and not just affordable tuition, but quite expensive tuition. And so that selects the students who are entering those schools, and it makes comparisons to public schools very difficult; which isn't to criticize those schools at all, but just to understand that things that might work in one setting might not easily translate into other settings.

There are, you said 50 alpha schools. I believe there are about 100,000 American schools right now. So we're talking about a very small model that's gotten a great deal of attention, understandably, but we should be cautious before thinking that it could easily expand and change the massive system of American public education, but we should also keep an open mind about what schools can learn from these new innovations.

LO: Matt, you've recently written about a study involving middle school students in Tennessee being offered AI tutors. What were the findings of that two-year study?

MB: So this was a study that looked at a program called Khan Academy, which some listeners may have heard of. It's designed, it often, it's known especially for its videos, but the program in schools is designed to help students walk students through math problems and give them math practice for students who are struggling. And so researchers looked at schools that introduced during the school day Khan Academy versus schools that didn't, or grade levels that didn't introduce Khan Academy, and were using other digital products or a small group teacher-led instruction.

And the other thing about Khan Academy is it recently introduced an AI chat bot known as Khanmigo, where students, if they were struggling, if they couldn't get the math problem right and the explanations in Khan Academy were not helping them, they could ask the chat bot. And unlike other AIs, it wouldn't give them the answer. So if the student were, if it was like, if the student was solving a problem and they're like, tell me the answer, the chat bot was programmed not to do so, but say, ‘well, let me help you walk through it. Here's the first step. Tell me the second step.’ And so the hope was that students would turn to this chat bot for help understanding the tricky math problems.

So the study had two very salient findings: Khan Academy was valuable. When students were practicing online with this program, they were learning more math than the other group of students that weren't on Khan Academy. ‘Was it because of the AI though,’ the researchers asked? Well, almost definitely not, because the students just weren't using the AI chat bot Khanmigo. The vast majority of days when they were on Khan Academy, they weren't engaging at all with Khanmigo. And then when the students were engaging with Khanmigo, they often weren't terribly productive. They were saying things like, ‘just tell me the answer already,’ which Khanmigo wasn't supposed to do and didn't, but that the student would get frustrated. Or there was one example in the study where the student just said to Khanmigo, ‘say a joke.’ And so there really wasn't any indication that the students were productively engaging with the AI chat bot in the way that designers and the educators were hoping.

NC: Could it be that EdTech initiatives like Khan Academy have been around for a while and have gained more acceptance while AI in the tutoring space is very new and it will take time for students to get engaged? Or do you think students will just always prefer human interaction when it comes to tutoring?

MB: I think the answer could be yes and yes to what you asked. That yes, there could be a learning curve for students and a learning curve for the developers of these products to make them more accessible and friendly and easy to use. And I know that Khan Academy has redesigned their program a bit to more seamlessly integrate Khanmigo into Khan Academy. So it's more salient and the students are more likely to use it. I don't think we have data yet on whether that's happening. And yet at the same time, we know from other evidence that human-to-human tutoring, teacher-to-student tutoring—doesn’t have to be a licensed teacher, but just like a skilled person, a skilled adult tutoring a student—is one of the most effective things in education. And people have been trying to automate that technology for a long time for understandable reasons, because we can't afford to give every single student all 50 million students their own personal tutor. Like they're just not enough humans, adults in schools to do that.

But what's striking is that that has not succeeded at least at replacing the human value of education. And so far, I don't think we're seeing AI chat bots succeed because part of educating young people especially goes beyond delivering the content. It's not just telling them how to do the algebra. It's motivating to care to do the algebra. Humans are good at that. And that's a skillset that teachers have. And teachers will tell you that's really hard. AI chat bots don't seem to be terribly good at that.

LO: Well, I'm gonna worry if some of these AI companies start hiring comedians as well as philosophers so that they can have jokes for these students. It's quite a world.

Also, another Chalk Beat article talks about Anthropic unveils Claude for teachers joining OpenAI and Google in the race to dominate the classroom AI. It's quite a competitive process. And the article talks about Newark Public Schools, AI tutoring chat bots, AI literacy screeners, and AI cameras across schools. Now, not connected with schools, but we've also been seeing reports about a few AI models, a few, going rogue in when they're doing these experimental cybersecurity testing, they've actually hacked a third party. It's limited, the companies are putting in safety parameters, but are you hearing concerns from parents or teachers about AI in the classroom, whether it's privacy or other kinds of issues regarding AI?

MB: I think there are a lot of concerns about AI broadly about how it's affecting society. And I think we need to be very careful and grounded in making sure we're talking about what is actually happening in schools versus the fears. But people are fearful and perhaps for understandable reasons because this is a new remarkably powerful but also largely untested technology that is changing our society and schools, maybe for the better, maybe for worse, maybe both. And it's very understandable that people have a lot of questions and a degree of skepticism about what, if any role that should be playing in American schools.

LO: So what is the best role for AI in American schools? What's the positive? We've talked about some of the potential negatives, but what is the positive?

MB: Well, I can't, as a journalist, I don't wanna say what the best role is, but I think what we have seen, there is potential and where I've seen people excited in education is AI helping teachers. Teachers generate a whole lot of text, and AI is really good at generating text. Teachers are writing notes home, they're creating rubrics, they're writing lesson plans, they're doing worksheets, they're giving students written feedback on their essays. And AI, I've talked to teachers who say, AI has helped them with that. It's made their jobs easier that they can have AI say, write a first draft of a rubric, help brainstorm a lesson plan, write the first draft of an email they wanna send out to parents. I think there are real risks to that. And I think there are risks to any professional who uses AI to do some of the work because sometimes doing the work is part of the work, in that like if you're writing a note home, you are thinking more deeply about how you wanna communicate to parents if you're a teacher. And if you're outsourcing that to AI or at least part of that to AI, there's a risk and there's a risk that teachers might notice that or there's a risk that it will design a lesson plan that isn't very good or that you're not as good at implementing because you didn't design it.

I think teachers I've talked to have tried to be really thoughtful about balancing that and thinking, okay, how can I have AI help but not do the heavy lifting in a way that is affecting me negatively? And it's also important to understand that, I just did a story showing that there's been a large rise in teachers leaving the classroom in the last several years. It's coming back down now but it had been very high for a few years. And so teachers are stressed out in many cases. If there are things that can make their jobs easier without sacrificing the quality of their teaching, that could be a big win.

NC: That's a great point. So the AI may help with the structure or the template but the teacher who knows the student who picks up maybe on the visual cues or the emotional cues or has a varied amount of emotional intelligence going into processing what that child needs, how they can use that template and then personalize it a little bit with a human touch, if you will. That sounds like potentially could be a useful tool.

Parents in California raised concerns after the Chico Unified School District Board reportedly approved a $1.9 million contract for new security cameras that have AI software capabilities. Since the system is on a cloud and footage is stored for 30 days, parents have also had some privacy concerns. With public funds being used for services controlled by private companies, what are you hearing about how parents and students are feeling about whose interests are being best served?

MB: I do think from a feeling perspective, there are a vocal contingent of parents and students who are concerned, who are concerned with the role of companies and vendors in American education and are worried about the value of agreement school districts might be making with companies and the role of AI and whether that's transparent and what exactly the role will play and what that will mean for the privacy of students and how exactly AI, what role AI will play in this. And I think those are fair concerns.

It's also worth keeping in mind that business in education has always been there and it must be there. Before AI and computers and Chromebooks, there were textbooks, and textbook companies sold textbooks and there were curriculums and there are companies that sold chalkboards and that sold chalk. And so it's not a new thing and there have been concerns raised over the years about companies in American education. So I think we should, I think it's fair for folks to ask questions. It's good for folks to ask questions. I also think we should make sure to understand this is not an entirely new phenomenon of a business or tech interest in education.

LO: You know Matt, we've talked about AI in the classroom and to what extent it supports learning versus replacing thinking. It's interesting because in Beijing, AI is mandatory and kids are not only actively being introduced to it, they are creating with it, learning about intelligent agents and algorithms. Is there any data internationally or nationally about how AI technology is improving teaching and learning?

MB: There is some data, but overall it is very nascent and there are not decisive conclusions to be drawn at this point. I think I would say there are two pieces of evidence that I think are quite important though. One, which we've already spoken about is that when students are given chatbots to help them learn, they often don't use those AI chatbots. And so that doesn't mean that they won't have value or they won't have value for some students, but it does mean that there's not just like, oh my gosh, students are flocking to AI to learn more.

And then there's some evidence, and I believe there was a recent study from China that shows if students, and also just like, this isn't a data point, but if you talk to any high school teacher or perhaps even middle school teacher, you will hear this, that if a student can outsource their thinking to AI and not do the work themselves, they will. Why? Because they're adolescents. And also some of us might do that the same way as adults, I could add, so it's not just adolescents who would take the easy way out. And so there is some reason to be concerned that without being very careful, AI can be used to outsource learning and avoid learning while completing homework assignments or other assignments. And that's why a lot of teachers are changing their approaches to try to prevent students from using AI.

LO: It raises the whole question of critical thinking skills, which is what college has always offered. And if you outsource some of that learning, you're not really using your critical thinking skills. Is that what you're saying?

MB: Well, I would say it a little differently actually, because critical thinking actually is not just an abstract skill, it's a skill that is based on your knowledge of a topic. If you know a lot about something, I'm an education reporter, so I know a lot about education. I can think very critically about education. I can analyze things, I can synthesize things, I can make my own conclusions about things because I know a lot about American education. But there are a lot of other things that I don't know much about. I don't know much about education in China. I don't know much about American foreign policy as it relates to China, for instance. I'm not a very good critical thinker when it comes to that.

And so there have been some people who say, "Oh gosh, you know, in the age of AI, we don't need to teach kids knowledge because they can just use AI to find the knowledge." And they said this about Google, and they said this about encyclopedias. You don't need to know things, you just need to be able to look them up and think critically. But it turns out having a deep knowledge gives you fluency that allows you to quickly analyze, synthesize, and think critically, that you can't do if you're just Googling something for the first time, or just using AI to understand something for the first time.

So I would say that if students are outsourcing their knowledge acquisition to AI, that could be dangerous and risky because that will reduce what they know, which will mean they will have a harder time thinking critically about a range of subjects.

NC: In the Chalkbeat article, Newark's new workforce program wants young adults to master AI before the job market demands it, it discusses how trainees are using artificial intelligence to prepare for careers, for example, with Dell's Gen AI mock interview tool to generate feedback about how people can improve their interviews. What do we know about how AI is impacting the job prospects of young college graduates?

MB: So this is a very interesting area because it seems like a very simple question: Is it harder to get a job now because of AI if you just graduated college? And there's actually a fierce debate among economists about this very simple question. What we do know is it is pretty hard to get a job right now if you're a young college graduate. The question is, is it because of AI? My reading of the evidence is no, that actually the job market is relatively weak for young people of a variety of education levels and it's not just those with a college degree. And there's also a recent paper that shows that actually some of the weak job market has had to do with remote work and a harder time of young people getting their foot in the door in that context. And so if you're a young person who's struggling to get a job, it doesn't seem like you're alone. It's not clear and my reading of the evidence is it's not likely that that is because of AI, but I think there is some real uncertainty about that.

LO: Just going back to the notion of AI and how it should be regulated, we know that new European Union regulations require companies to clearly label AI generated content and non-compliance carries penalties. Some states have already adopted some AI transparency laws, but do you think we'll see more of those laws coming in future years?

MB: I don't know and I'll pivot to my area of expertise, which is schools and say like, how will schools approach that sort of regulation? And what we know is that schools in America are highly decentralized. We do not have a national government that is telling schools exactly what they should do, how they should be using AI. The Trump administration has put out various guidance, documents and declarations about AI and EdTech, but none of it is, I don't believe any of it is legally binding. So schools and districts and to some extent states, which do have more of a role in schools, have a pretty free hand about how they use or don't use AI tools.

And I think that has been a blessing and a curse for American education in that you're going to see a lot of different use cases or non-use cases and some of it is going to be good, some of it is going to be bad and schools and districts are largely figuring it out and also teachers are largely figuring it out on their own.

LO: The recent landmark case in which Meta settled in 47 states agreeing to pay more than $17 billion raises questions about the impact on other tech companies. Now, of course, this case revolved around claims regarding social media and endangering children with addictive platforms, allegations which of course Meta has denied, but Meta has agreed to some new safety features for children and teens on Facebook and Instagram. And although social media is a very different platform than what we're talking about AI in the classroom, do you think that the notion of student tutoring, offering a personalized learning experience, those sorts of things will be impacted by this Meta case?

MB: I don't know that they will be impacted directly, but what I do think is that these sorts of cases and these sorts of concerns have created skepticism among the public and parents about how tech affects young people. And that drives skepticism beyond the social media companies that are at the heart of these cases and those concerns. And I think that sort of skepticism has fed into people's skepticism about tech companies promoting AI in schools and the role of AI in schools.

NC: Matt, what motivational speech would you give at a commencement?

MB: I guess I would say, and I think about this with my own work in the middle of my own career in an industry that is in flux because of AI: Journalism. This is a scary moment, but it's also a really exciting moment. And it's really interesting to watch this evolve and watch how society is changing. And again, that's not to downplay the scary aspect of it, and not to assume that everything is going to end well.

But if I were a young person, I would lean into what excites me, whether it has anything to do with AI or not. I also would say that I don't know that I would totally tear up whatever career plans I already had because of AI. There are a lot of people making confident predictions and saying, you should do this or you shouldn't do this, or you must learn this, or you can't learn this. And I still go back to that there is a great deal of uncertainty. So radically altering your life plans because there is more uncertainty, but without confidence about what that's going to look like may not be a wise idea.

LO: Thank you, Matt Barnum. You can read more articles about AI and education at chalkbeat.org. That's C-H-A-L-K (as in teachers’ chalk), chalkbeat.org. Matt Barnum, editor and columnist for Chalkbeat Ideas. Thank you for joining us.

MB: Thank you for having me and for the rich conversation. I really appreciate it.

Thank you for joining us for Navigating Our Financial Times. Our producer is Anne Simmons. Our audio engineer is Gabe Stultz and our production engineer is Jim Blackwood. This program is not intended to provide investment, tax or financial advice. You should consult your own tax, legal and accounting advisors before engaging in any transaction. The views and opinions expressed by guests are their own and do not imply an endorsement or recommendation or necessarily reflect the views of AZPM.

Support comes from Banner Health, a nonprofit healthcare system with an integrative health plan committed to giving back to Arizona. Last year, Banner reinvested $1.1 billion into the Arizona community. Learn more at bannerhealth.com. This has been an original production of AZPM News.